Vaccination of mice with either third-stage Ancylostoma caninum infective hookworm larvae (L3) or alum-precipitated recombinant Ancylostoma secreted protein 1 from A. caninum (Ac-ASP-1) results in protection against hookworm challenge infections. Vaccine protection is manifested by reductions in lung hookworm burdens at 48 h postchallenge. Mice actively immunized 4 times with Ac-ASP-1 also exhibited reductions in hookworm burden in the muscles. Hookworm burden reductions from Ac-ASP-1 immunization were associated with elevations in all immunoglobulin subclasses, with the greatest rise observed in host IgG1 and IgG2b. The addition of a fourth immunization resulted in even higher levels of IgG and IgE. In contrast, L3-vaccinated mice exhibited marked elevations in IgG1 and IgM, including anti—Ac-ASP-1 IgM antibody. Passive immunization with pooled sera from recombinant Ac-ASP-1—vaccinated mice also resulted in lung hookworm burden reductions. It is hypothesized that recombinant Ac-ASP-1 vaccinations elicit antibody that interferes with parasite larval migration.
Hookworm infection is a major cause of anemia and malnutrition in underdeveloped regions of the tropics and sub-tropics. An estimated 194 million cases of hookworm infection occur in China alone . For almost a century, mass chemotherapy with anthelminthic drugs has been the major weapon used to control hookworm infection. This approach has failed in some endemic areas because of either high rates of post-treatment reinfection  or outright anthelminthic drug failure . As an alternative to chemotherapy with anthelminthic drugs, we have suggested that it might be possible to vaccinate individuals against hookworm infection  Our focus for vaccine targets has been infective third-stage larvae (L3) of the dog hookworm Ancylostoma caninum. The rationale for choosing A. caninum L3 is based on the older observations that these parasites are antigenic and can protect laboratory animals against hookworm-challenge infections when administered in small doses [5, 6] or in large doses when they are previously attenuated by ionizing radiation . To characterize some of the major antigens of the infective larval stages of A. caninum, their major excretory/secretory (ES) proteins were isolated, and their corresponding cDNAs were cloned. Among them are a metalloprotease  and 2 predominant proteins of unknown function known as the A. caninum secreted proteins (Ac-ASPs) . Ac-ASP-1 (45 kDa) is the major ES protein released by infective A. caninum L3 after activation under host-stimulatory conditions , and Ac-ASP-2 (24 kDa) is a less abundant constituent . Ac-ASP-1 may be a heterodimorphic repeat of the monomorphic Ac-ASP-2. Well-conserved orthologues of the asp cDNAs were recently identified and cloned from the human hookworms Ancylostoma duodenale (Ad-asp-1) and Necator americanus (Na-asp-1) .
To determine whether recombinant ASP proteins (expressed in Escherichia coli) are effective vaccine candidates in laboratory animals, we employed a mouse model of A. caninum hookworm infection first developed by Kerr in 1936  and later by Behnke . Our studies rely on the extraintestinal migrations of A. caninum L3 after oral entry . Briefly, A. caninum L3 introduced into the gastrointestinal tract via gavage tube will penetrate the gut and enter the lungs beginning 24 h after administration. By 48 h, the L3 enter the lungs in maximal numbers, where they remain for ∼12 h. L3 migrating in the lungs elicit pulmonary hemorrhage, a condition sometimes referred to as verminous pneumonitis . From the lungs, some L3 reenter the gastrointestinal tract (presumably by ascending the bronchial respiratory tree), while the remainder enter the muscles (presumably by gaining access to the systemic efferent circulation). L3 can enter the muscle as early as 48 h after infection but arrive in large numbers beginning 1 week after infection. Over the ensuing 2–3 weeks (3–4 weeks after infection), L3 continue to accumulate in the muscles. A. caninum L3 do not develop into adult hookworms in mice.
Kerr first noted that mice could be immunized with live L3; these vaccinated mice exhibit reductions in hookworm burden and do not develop severe verminous pneumonitis . We reported previously that mice vaccinated with alum-precipitated recombinant Ac-ASP-1 also exhibit significant reductions of A. caninum L3 in the lungs after larval challenge, compared with negative controls . This observation is now further extended to include the muscles and gastrointestinal tract at later times in the infection. We also report that hookworm burden reduction is antibody dependent and may operate by blocking L3 entry into the lungs.
Materials and Methods
Mouse immunizations, hookworm challenge infections, and evaluation of vaccine protection
Recombinant Ac-ASP-1 (a 1-kb EcoRI/Xho1 fragment containing the 3′ end and poly[A] tail of Ac—asp-1 cDNA expressed in E. coli by use of the expression vector pET 28 [Novagen]) containing polyhistidine and T7 tags and amino acids 96–424 of Ac-ASP-1  was purified and precipitated with aluminum potassium sulfate dodecahydrate and sodium bicarbonate, as described elsewhere . Briefly, the sodium dodecyl sulfate-solubilized immunogen is precipitated with an aqueous solution of the aluminum salt under alkaline conditions prior to centrifugation and washing. Outbred 5-week-old cesarean-derived male mice were immunized with 20 µg of alum-precipitated recombinant Ac-ASP-1 by intraperitoneal injection, followed by either 2 or 3 subsequent boosts . Age-matched unimmunized mice served as negative controls. For comparative experiments, mice were also “immunized” with L3 of A. caninum by receiving 3 consecutive infectious doses of 500 L3 [14–17]. The L3 were obtained by coproculture of feces from A. caninum—infected beagle pups . All immunized and control mice were challenged with 500 L3 per os, after the final immunization, prior to euthanasia with Metofane (Mallinckrodt Veterinary, Mundelein, IL).
Hookworm burden reductions were measured by counting the number of L3 recovered from the lungs, gastrointestinal tract, and skeletal muscles of immunized and unimmunized control mice at 48 h and 1, 2, and 3 weeks after challenge. Alternatively, L3 were recovered from gastrointestinal tract, lungs, and liver of unimmunized naive mice at 12, 24, 36, 48, and 72 h after infection. Lung L3 were recovered from minced lung and counted with the aid of a dissecting microscope as described elsewhere . L3 were also recovered from the liver by use of the same technique. The L3 from the gastrointestinal tract were released from the tissues by mechanical shearing with scissors followed by chemical digestion (0.7% HCl in PBS containing 0.5% pepsin) for 2 h of constant shaking at 37°C. The L3 were recovered by placing the digested gastrointestinal tract in a Baermann apparatus for 3–6 h and were then counted. The L3 were recovered from mouse skeletal muscle dissected from the carcass in the same manner. The statistical significance of hookworm burden reductions from vaccinated versus control mice was determined by use of the Mann-Whitney non-parametric test.
Antigen-dependent serum isotyping
Anti—Ac-ASP-1—specific IgG and IgM antibodies in mouse sera were measured by ELISA as described elsewhere . To measure anti-L3-specific antibodies, the procedure was modified by substituting recombinant Ac-ASP-1 with 0.4 µg of soluble A. caninum L3 antigen. Antigen-specific IgA, IgM, and IgG subclass antibody measurements were made by use of the ImmunoPure Monoclonal Antibody Isotyping Kit (product 37501) from Pierce (Rockford, IL) as recommended by the manufacturer.
Total serum IgE was measured separately. Briefly, a Dynatech (Chantilly, VA) microtiter plate was coated with 0.3 µg of sheep anti-mouse IgE (Binding Site, Birmingham, UK) in borate saline (0.17 M sodium borate, pH 8.0, 0.12 M sodium chloride) overnight at 4°C. The plate was washed twice with 0.05% Tween 20 in PBS and then again in PBS alone. Mouse serum samples were added at dilutions beginning as low as 1 : 20 and allowed to react overnight at 4°C. An IgE standard from Pharmingen (San Diego, CA) served as a positive control. The solution was discarded and washed, as described earlier, before biotin-conjugated rat anti-mouse IgE (BioSource International, Camarillo, CA) was added. The mixture was incubated for 1 h at 37°C before streptavidin-conjugated horseradish peroxidase (Zymed, South San Francisco) was added and incubated for an additional 45 min at 37°C. After washing, the microtiter plate was read at 405 nm absorbance (ImmunoPure ABTS tablets; Pierce).
For measurement of anti—Ac-ASP-1 IgE, the plate was coated with recombinant Ac-ASP-1 as described earlier. The plate was washed twice with 0.05% Tween 20 in PBS and then with PBS alone. The wells were blocked with 1% bovine serum albumin in borate saline. Sera from unimmunized mice or mice vaccinated with either recombinant Ac-ASP-1 (immunized both 3 and 4 times) or A. caninum L3 were added to the wells overnight at 4°C. Antisera dilutions ranged from 1 : 20 to 1 : 160. After washing, biotin-conjugated rat anti-mouse IgE (Pharmingen) was added at a 1 : 1000 dilution, and the plate was incubated at 37°C for 1 h. Peroxidase-conjugated goat mouse anti-biotin (Boehringer, Mannheim, Germany) in blocking buffer was subsequently added at a 1 : 400 dilution. The plate was incubated at 37°C for 45 min prior to washing and incubation with ABTS substrate. The microtiter plate was read at 405 nm.
Passive antibody transfer
Eighty microliters of pooled sera from either recombinant Ac-ASP-1 vaccinated mice (both 3 and 4 immunizations), L3 vaccinated mice, or unimmunized (naive) mice was diluted with 2× volume PBS for a total volume of 240 µL. The diluted immune or naive sera were administered by intraperitoneal injection (ip) either at the time of L3 infection or at 24 h after infection. The mice were killed 48 h after L3 challenge, and the number of L3 recovered from the lungs and gastrointestinal tract was counted, as described earlier. The statistical significance of hookworm burden reductions from vaccinated versus control mice was determined by use of the Mann-Whitney nonparametric test.
Path of A. caninum L3 migration
The recovery of L3 from the gastrointestinal tract, liver, and lungs of unimmunized mice at 12, 24, 36, 48, and 72 h after infection is shown in figure 1. The L3 begin to exit the gastrointestinal tract within 12 h after infection, when they can start to be recovered from the liver. The greatest number of L3 are recovered from the liver by 24 h after infection. By 36 h after infection, most of the L3 are no longer found in the liver but instead have started to enter the lungs. The greatest number of L3 are recovered from the lungs by 48 h after infection. From there they begin exiting the lungs and arrive in the muscles, where they accumulate over the next 1–3 weeks (table 1).
Hookworm-burden reductions: vaccine protection
As shown in table 1, mice vaccinated ip with alum-precipitated recombinant Ac-ASP-1 and then boosted twice (mice vaccinated 3 times) showed a significant reduction of hookworms in the lungs at 48 h after challenge (57%), compared with findings in control mice. Mice vaccinated 3 times did not exhibit significant reductions in L3 in the muscles at 1, 2, and 3 weeks after challenge. Of interest, there was an increase of L3 in the gastrointestinal tract of mice vaccinated 3 times at 48 h after challenge infection. Mice vaccinated 4 times with recombinant Ac-ASP-1 also exhibited hookworm burden reductions in the lungs. As shown in table 2, in 2 different experiments the mice vaccinated 4 times exhibited a 63% reduction of L3 in the lungs at 48 h after challenge compared with unimmunized mice. In addition, there was a significant hookworm burden reduction in the gastrointestinal tract and muscles when mice were given an additional boost of recombinant Ac-ASP-1. Hookworm burden reductions in these 2 organs were apparent at 1, 2, and 3 weeks after challenge.
Humoral antibody responses
As shown in figure 2, mice vaccinated 3 times with alum-precipitated recombinant Ac-ASP-1 exhibited elevations in all anti—Ac-ASP-1 immunoglobulin classes and subclasses. Levels of immunoglobulin G1 showed the greatest elevation with measurable anti—Ac-ASP-1 specific antibodies at a 1 : 100,000 dilution; IgG2b levels were also significantly elevated. For comparison, mice vaccinated 3 times with hookworm L3 also exhibited demonstrable increases in anti-L3 IgG1 levels but with a significant elevation in anti-L3 IgM antibody titer (figure 3). The total serum IgE levels were elevated in both groups of vaccinated mice. The serum IgE concentration in unimmunized mice was 3 ng/mL but was elevated to 24 ng/mL in recombinant Ac-ASP-1 vaccinated mice and 448 ng/mL in L3 vaccinated mice, corresponding to 8- and 149-fold increases, respectively (data not shown). As shown in figure 4, there were diferences in anti—Ac-ASP-1 antibody responses between mice vaccinated with Ac-ASP-1 3 and 4 times and with L3. Mice vaccinated with L3 exhibit prominent IgM anti—Ac-ASP-1 antibody responses. Levels of the other classes and subclasses of immunoglobulins were not significantly elevated relative to those seen in unimmunized mice. In contrast, mice vaccinated with Ac-ASP-1 exhibit prominent IgG and IgE anti—Ac-ASP-1 antibody responses. The amount of anti—Ac-ASP-1 IgG and IgE antibody increases with a fourth immunization (third boost).
Passive transfer of immunity
Diluted pooled sera from mice immunized against Ac-ASP-1 3 and 4 times were administered ip to mice either at the time of oral hookworm infection or 24 h after infection. Hookworm burden reductions were evaluated by counting the number of L3 recovered from the lungs. As shown in table 3, passive transfer of either the anti-L3 antiserum or both anti—Ac-ASP-1 antisera (from mice immunized either 3 or 4 times) resulted in significant reductions in lung hookworm burden relative to passive transfer of serum obtained from unimmunized mice (54%–63%). These antisera were administered ip at the same time that the challenge dose of A. caninum L3 larvae was administered orally. In contrast to the reduction of L3 in the lungs of passively immunized mice, there was an increase of L3 in their gastrointestinal tract. Anti—Ac-ASP-1 antisera elicited a greater increase than did anti-L3 antiserum.
Table 3 also shows the result of passive antibody transfer when the timing of antiserum administration is delayed 24 h. for these experiments, the antisera were administered ip 24 h after infection. As shown in table 3, passive immunization with either anti-L3 or anti—Ac-ASP-1 antiserum resulted in lung hookworm burden reductions (60%–68%) similar to those described earlier. However, passive immunization with anti-Ac-ASP-1 antiserum resulted in a large recovery of L3 from the gastrointestinal tract. It was estimated that there were 464% and 482% increases in gut L3 after passive immunization by use of pooled antiserum from mice immunized 3 or 4 times, respectively.
We showed previously that active immunization with alumprecipitated recombinant Ac-ASP-1 results in reductions in the number of hookworm L3 recovered from the lungs at 48 h after challenge compared with immunization in mice injected with either saline  or alum (unpublished observation). Reductions in worm burden are typically determined by recovering L3 that migrate out of the minced tissues. Here we extend our observations by examining other organs (gastrointestinal tract and muscles) in mice vaccinated either 3 or 4 times with alumprecipitated recombinant Ac-ASP-1. The organs examined were those from which L3 were recovered. Evidence is provided that the hookworm burden reductions resulting from Ac-ASP-1 vaccination are antibody mediated.
We determined that after oral administration of A. caninum L3 to mice, the L3 exit the gut within 12 h after administration and enter the liver and then the lungs at ∼24 and 48 h after administration, respectively. From the lungs, the L3 begin to accumulate in significant numbers in the muscle over the ensuing 1–3 weeks. In some instances, the L3 return to the gastrointestinal tract. We determined that mice actively immunized 4 times with alum-precipitated recombinant Ac-ASP-1 (1 primary immunization and 3 boosts) exhibit significant reductions of hookworm L3 recovered from all target organs. Mice immunized only 3 times, however, exhibit reductions in the lungs only.
Mice responded to immunization with alum-precipitated recombinant Ac-ASP-1 by exhibiting elevated levels of all anti—Ac-ASP-1 antibody classes and subclasses examined, but IgG1 showed the greatest elevation, followed by IgG2b. Mice that received an extra boost of Ac-ASP-1 responded with increased levels of anti-Ac-ASP-1 IgG and IgE. We are investigating whether the increase in anti—Ac-ASP-1—specific IgG and IgE may account for the greater reduction in gastrointestinal tract and muscle hookworm burdens. Mice vaccinated with L3 exhibit significant increases in parasite-specific IgM and IgG1. These elevations were similar to those reported elsewhere from mice immunized with Strongyloides stercoralis L3 . IgM antibody has been noted to be associated with protection in Strongyloides-immunized mice . The anti—Ac-ASP-1 antibody responses in A. caninum L3-vaccinated mice are predominantly of the IgM class.
The finding that hookworm lung burden reductions could be passively transferred with antibody confirms the importance of ASP-1—specific antibody in mediating reductions in lung hookworm burdens. We do not know why passive immunization with anti—Ac-ASP-1 antibodies will result in an increase of A. caninum L3 in the gastrointestinal tract, particularly when antibody administration is delayed 24 h after oral infection. One explanation for the increase in gut L3 when passive immunization occurs at 24 h after infection is related to the location of L3 in the mouse at this time. At 24 h after infection we were able to recover ∼two-thirds of the total inoculum of L3. Of these, about 200 L3 were recovered from the gastrointestinal tract and almost 100 from the liver (figure 1). Therefore, the ip delivery of anti—Ac-ASP-1 antibodies may either trap L3 in the gastrointestinal tract and prevent their exit or signal the L3 in the liver to return to the gut. Although we did not observe accumulation in L3 recovered from the gut at 48 h in mice actively immunized 4 times, we have sometimes observed this phenomenon in our inbred mouse strains (unpublished observation). Possibly the variation in L3 recovery from the intestine reflects variable rates of larval excretion in the feces. We did not attempt to recover L3 from mouse feces.
At this time it is not clear how host antibodies would function to block parasite migration; in this case, how antibodies would prevent larval exit from the gut or force their return from the liver. There is precedent for murine host immune responses blocking parasite migration, such as the T cell responses that retard the migration of challenge schistosome cercariae in the lungs . However, we did not find prominent host cellular inflammatory responses in our recombinant Ac-ASP-1 immunized mice that could account for this mechanism . The preliminary observation that ASP-1 may be an amphidial protein in the plant pathogenic nematode Meloidigyne incognita (R. Hussey, University of Georgia, personal communication) suggests the possibility that anti—ASP-1 antibodies could affect certain amphidial functions (e.g., chemosensation) associated with larval migration. Studies are in progress to evaluate whether anti—ASP-1 antibodies (especially antibodies of the IgE class) affect hookworm migration in vivo through tracer studies with 35S- radiolabeled L3. Radiotracer studies will also help us to determine whether significant numbers of larvae remain trapped in the lungs or other tissues.